Instrument air vs. Electricity: Reliability issue

[shahyar]

There is a discussion with my colleagues: Which one in more reliable: Instrument air (IA) or electricity?
The question raised when we were talking about the fail position of a switch valve which actuates by IA (energized by a solenoid).
Is there more chance of loosing IA or electricity?
Thanks for your answers.

 

[ash9144]

If your IA is generated by an electric compressor then you are more likely to lose IA. Since your likelihood would be equal to likelihood of losing electricity plus compressor failure plus whatever other failure mechanisms you may have.

If you have instrument air from a different source (or different electric supply) then it may be different.

 

[zdas04]

Not sure I completely agree with that. It really depends on how you define a "loss" and what are the consequences of a loss. If a blip in line power can drop latching relay's and reposition valves, that is a problem that you don't have if your IA has an adequately sized receiver. A short duration electrical outage may stop the compressor from running, but the stored air will operate many valves for a while, maybe long enough.

To me the big drawback of electrical system is that it is difficult and expensive to store power.

For this kind of analysis I generally list everything that I can imagine that can go wrong. Then I put a column for electrical and a column for pneumatic. I grade the risk of each line happening in this facility for electric or pneumatic (usually on a 0-10 scale , with high numbers being higher risk), then add up the columns. Low number wins. Sometimes it is electric that wins, sometimes it is pneumatic that wins. In my experience there is not a universal answer.

David

 

[quark]

I agree with David fully. Since, we can store compressed air, the pneumatic valves can remain activated provided solenoids have electrical power. This power requiremen will be less than that of electrically actuated valves.

On the other hand, if the solenoids are of manifold type (compact block type), one solenoid leaking straight into the exhaust manifold may disturb other valves as well.

These conditions are for normal operation.

If valves are good (that coils are not damaged, diaphragms are free to move etc.) there shouldn't be any problem with fail safe condition. Either of the valves get actuated to fail safe position.

 

[europipe]

Also take care of hazard explosion area's with
electric steering.

 

[25362]

Instrumentation UPS back up is quite common in oil refineries.

 

[shahyar]

Thanks for all your responses.
I just feel when we are talking a "probability" of a loss, we should not bring "consequence" of a loss in the evaluation. They are two independent factors.
I agree to evaluate a risk, both of them should be addressed.
Thanks again.

 

[zdas04]

That is why I list the events that can happen, and then evaluate the consequences of each. Two different steps in the same analysis.

It is pretty dumb to treat them each in their own vacuum. You should pay more attention to a choice that has a low probabality, but a huge consequence than you would pay to something that is very likely, but has a very low consequence.

David